Evaluating the performance of Metrology Wells

1
Evaluating the performance of Metrology Wells
Its a Metrology Well, not a Dry-Well!
2
Metrology well design

• Addresses every performance category in EA-10/13
• New control techniques
• stability to 0.005K
• Dual zone control
• Approaches bath level performance
• Built in reference thermometry
• Accepts ITS-90 constants
• Comes with traceable accredited calibration

3
How do you determine dry-well accuracy?
Need to know

• How the dry well will be used
• Important sources of error
• How dry well manufacturers write specifications

4
How you use a dry-well greatly affects
performance!

• Temperature range
• in general errors increase further from ambient
• Will dry-well temperature be measured using
  external reference or internal control sensor and
  display
• each method is valid, but an external reference
  will generally provide better uncertainties

5
How you use a dry-well greatly affects
performance!

• Metrology wells and dry wells are used for
  comparison calibration
• Comparison calibrations require thermal
  equilibrium and consistency
• without thermal equilibrium cannot make
  measurements
• equilibrium requires stability
• consistency allows comparisons to have meaning
  over time and between different tests
• good consistency requires similar loading, low
  drift, good handling practices and verification

6
Errors depend upon mode of use

• External Reference
• Axial Uniformity
• Radial Uniformity
• Stem conduction
• Loading Effect
• Stability
• Reference temperature measurement
• reference probe
• reference readout
• hysteresis

• Internal control sensor and
• calibrated display
• Axial Uniformity
• Radial Uniformity
• Stem conduction
• Loading effect
• Stability
• Reference temperature measurement
• sensor and display drift
• hysteresis
• sensor calibration

7
Estimate of significance
Note different scaling
8
Axial uniformity
Temperature difference between top and bottom of
well

• Measurement zone is where axial uniformity is
  smallest
• EA10/13 requires 40mm
• Hart recommends 60mm
• Axial uniformity in measurement zone needs to be
  known to determine uncertainty
• Hart has special probe to measure this error
• Errors can be minimised by
• Dual zone control
• Ensuring sensor sits in measurement zone
• Aligning centres of reference and UUT

9
Radial uniformity
Temperature difference between holes and
reference at same depth

• Primarily a function of distance and material
  type
• External references can be closer to UUT
• Radial uniformity needs to be known to determine
  uncertainty
• Minimised by placing reference close to UUT

10
Stem effect
Heat conducted up the sheath of the probe

• Causes non equilibrium between sensor and source
• Function of size and type of material
• large diameter probes conduct more heat
• Al conducts more than Inconel
• Error minimised by deeper immersion
• Hart suggests 15-20 times probe diameter
• 8mm probe should have 120mm to 160mm depth
• Metrology wells have extra depth to minimise this
  error

11
Loading effect

• Number of probes will impact heat drawn from or
  into well
• Loading effect minimised by
• deeper immersion
• dual zone control

12
Temperature stability
0.181C 30 min 0.046C 15 min

• Stability is temperature variation with time
• Time should be specified
• EA10/13 recommends 30 minutes
• Should be stated with high confidence
• Use of TYPICAL to be avoided

0.017C 30 min 0.011C 15 min
13
Temperature stability
0.181C 30 min 0.046C 15 min

• Stability required to reach thermal equilibrium
• Probes need time to reach equilibrium with their
  surroundings
• Multiple measurements rarely instantaneous
• Stability error minimised by design
• Accurate control with good resolution

0.017C 30 min 0.011C 15 min
14
Reference temperature measurement
Sources of error to consider

• External Reference
• Probe calibration uncertainty
• Probe handling
• Probe characteristics
• Readout calibration uncertainty
• Readout resolution
• Measurement technique

• Internal control sensor and display
• Well calibration uncertainty and procedure
• Hysteresis
• Probe drift
• Control electronics calibration
• Control electronics resolution
• Control measurement technique

15
Estimate of significance - recap
Note different scaling
16
Total uncertainty reference probe
17
Total uncertainty internal reference
18
Comparison of modes of use

• All Metrology Wells have internal reference
  readout option
• With calibrated probe unmatched performance is
  achieved

19
Comparison
Compare cold Metrology Wells to dry wells
20
Comparison
Compare mid temperature Metrology Wells to dry
wells
21
Comparison
Compare high temperature Metrology Wells to dry
wells
22
EA Guideline

• EA-10/13 Guidelines on the Calibration of
  Temperature Block Calibrators
• ASTM guidelines are being written
• Harts Tom Wiandt on this committee
• Several NMIs have published or are revising
  their guidleines

23
EA Guideline

• One of first Guidelines available
• Approved by EA General Assembly in 2000
• Document available free of charge from
• www.european-accreditation.org
• Provides technical and procedural guidelines and
  a suggested method for calculating uncertainty

EA-10/13 is not a standard it is a suggested
guide to assessors and laboratories seeking
European Accreditation